PROJECT SUMMARY Breast cancer is the most common cancer in women and is the second leading cause of cancer mortality among women. Obesity is an important risk factor for breast cancer incidence among postmenopausal women. Recent evidence points to the role of central and visceral adiposity in the link between obesity and breast cancer, and to the differences in body fat distribution and the extent of its association with breast cancer by race/ethnicity. Estimation of a woman's breast cancer risk is an important tool used in primary breast cancer prevention efforts. Various models for predicting absolute risk of breast cancer have been developed for a number of racial/ethnic groups in several locations, but their ability to discriminate women who will eventually develop breast cancer from those who will not remains limited, with the area under receiver operating characteristic curve (AUC) generally ranging between 0.54-0.68. Efforts to include BMI in breast cancer risk models did not produce substantive improvement in the AUC. The effect of body fat distribution on predicted breast cancer risk mostly remains unexplored but has the potential to improve breast cancer risk prediction models. Thus, we propose to construct breast cancer risk models that would account for adiposity characteristics through the development of composite adiposity prediction scores based on anthropometry, and subsequent inclusion of these scores in breast cancer risk models. We will conduct these analyses among female participants of the Multiethnic Cohort, which includes five major racial/ethnic groups: African Americans, Japanese Americans, Latinos, Native Hawaiians and Whites. In Aim 1, in a cross-sectional study of 939 healthy women, we will examine the associations of imaging-based adiposity measures with clinically measured anthropometric characteristics; calibrate the self-reported anthropometric characteristics based on clinically measured anthropometry; and construct composite prediction models for each adiposity measure, using calibrated self-reported anthropometric characteristics within each racial/ethnic group. In Aim 2, using data on 51,336 female MEC participants, we will construct breast cancer risk models using different combinations of known risk factors for postmenopausal breast cancer and composite predicted adiposity measures from Aim 1. We will then evaluate predictive and discriminatory accuracy of these models, select the best-performing model within each ethnic group, and test these ethnic-specific models across racial/ethnic groups to examine the feasibility of a trans-ethnic breast cancer risk model. Breast cancer risk prediction models that include adiposity measures will have greater accuracy in predicting breast cancer, which would benefit prevention efforts by better informing patients of their risk level and guiding personalized prevention strategies across or within five different racial/ethnic populations.